Retrospective studies have demonstrated that anti-annexin V (anti-AnxV) antibodies are linked to miscarriage. Their predictive value is, however, unknown. We have carried out a prospective study to evaluate the relationship between anti-AnxV antibodies and the pregnancy outcome. A serum sample was taken from 1038 consecutive healthy women at the beginning of pregnancy. IgG and IgM anti-AnxV antibodies were measured by an ELISA method. The cutoff value was set at 5 units for both IgG and IgM. Out of 1038 women, 116 (11.4%) had a miscarriage by the 22nd week; 10 were lost to follow-up, 10 had an induced abortion, 6 had a preterm delivery, and 896 carried their pregnancy through to term. An adverse outcome of the pregnancy proved to be directly related to the number of previous miscarriages (P = .008) and the age of the woman (P = .002). IgG and IgM anti-AnxV were present in 25% and 27% of the women who miscarried, and in 23% and 28% of those who gave birth (mean antibody concentration IgG, 4.2 vs. 4.4 U/mL; IgM, 3.7 vs. 3.5 U/mL). IgG and IgM anticardiolipin and anti-beta(2)GPI, together with antinuclear, antithyroperoxidase, and antithyroglobulin antibodies, were also measured in the 116 sera of the women with miscarriage and in an equal number of women who gave birth. Their positivity or level proved not to be useful in discriminating between the risk of miscarriage and term delivery. This large-scale prospective study demonstrates that the presence of IgG and IgM anti-AnxV antibodies, when measured in healthy women, does not give a positive predictive lead towards the possibility of a miscarriage, and it is not useful in evaluating the risk of miscarriage at the beginning of pregnancy.